Conformational study of 3-substituted thietane 1-oxides. Lanthanide

David J. H. Smith, John D. Finlay, C. Richard Hall, and J. J. Uebel. J. Org. Chem. , 1979, 44 (26), pp 4757–4761 ... D. C. Dittmer , T. C. Sedergran...
0 downloads 0 Views 663KB Size
J . Org. Chem., Vol. 44, No. 26, 1979 4757

3-Substituted Thietane 1-Oxides aromatic),7.67-7.80 (m.1 H, aromatic);NMR (180 MHz, CDCl,) 6 1.80-2.04 (m, 2 H, Hi,*,anti to H3,?),2.06-2.26 (m, 2 H, H4,4, syn to H3,?),3.10-3.30 (m,6 H, H2,2',H3,3z,H5,5 syn to H3,3,), 3.40-3.64 (m, 2 H, H5,5anti to H3,,,),7.14-7.36 (m,6 H, arom), 7.46 (d of d, 2 H, J = '7, 2.5 Hz, H8,v).25 Anal. Calcd for C22112002:C, 83.51; H, 6.37. Found for 22: C, 83.59; H, 6.43. Fou:id for 23: C, 83.48; H, 6.42. Found for 24: C, 83.38; H, 6.38. Found for 25: C, 83.60; H, 6.23. (25) From the model of 25, the carbonyl group is almost perpendicular to the aryl ring. Consequently H8,8, protons are at higher field than in the other three dimers. I n 25, HS,5,protons anti to H3,? are equatorial and nearly in the plane of the aryl ring and consequently deshielded. The protons anti to H,,,. ,ire in pseudoaxial positions and are shielded by the carbonyl group.

Acknowledgment. We are indebted to the National Science Foundation (Grant CHE-05956) and the National Institutes of Health (Grant GM-15997) for financial support. We are also indebted to Dr. Donald L. Ward for the X-ray structure determinations. Registry No. 14, 69456-01-5; 15, 45998-98-9; 16, 71975-02-5; 17, 72028-55-8; 18, 71975-03-6; 19, 72028-56-9; 22, 72028-85-4; 23, 71975-04-7; 24, 71975-05-8; 25, 71975-06-9; 4,5-benzotropone, 444391-8. Supplementary Material Available: Positional and thermal parameters and their standard deviations for the X-ray structures of 18,22,23,24, and 25 (9 pages). Ordering information is given on any current masthead page.

Conformational Study of 3-Substituted Thietane 1-Oxides. Lanthanide Shift Reagent Approach David J. H. Smith,* John D. Finlay, and C. Richard Hall Department of Chemistry, T h e University of Leicester, Leicester, England LEI 7RH

J. J. Uebel* Department of Chemistry, T h e University of New Hampshire, Durham, New Hampshire 03824 Received M a y 29, 1979

The NMIt lanthanide-induced shifts (LIS) for a series of cis and trans 3-substituted thietane 1-oxides and 3,3-disubstitutedthietane 1-oxideswere measured and interpreted in terms of the ring conformations. The data suggested that ail cis 3-substituted thietane 1-oxides (R = CH,, t-Bu, and aryl) exist exclusively in the diequatorial conformation and that the two isomeric 3-methyl-3-phenylthietane 1-oxidesexist strongly, if not exclusively, in the equatorial oxygen conformation when they are bound to a shift reagent-Eu(dpm), or Yb(fod)$ The data for trans 3-substituted thietane 1-oxidessuggested that they all preferred the equatorial oxygen conformation (CH,. 86%;aryl, 75%; t-Bu, 65-7570) when they are bound to a shift reagent. Relatively little is known about the conformation of four-membered heterocycles such as thietane and its derivatives.' Pioneering work by Johnson2 and co-workers showed that the sulfinyl group of 3-substituted thietane 1-oxides prefers to occupy the equatorial position. In his work 3-tert-butyl- and 3-arylthietane 1-oxides were equilibrated to mixtures which were rich (85%)in the cis isomer. Arguments based on the chemical shift of H(3) were presented which suggested that the cis isomers were mainly in the diequatorial conformation and that the trans isomers were mainly in the equatorial R/axial 0 conformation. More recently one of us (D.J.H.S.) presented proton NMR and X-ray evidence which questioned the latter s ~ g g e s t i o n . Thus ~ there is some uncertainty about whether the sulfinyl group prefers the equatorial position more than 3-aryl and 3-alkyl substituents.

Results and Discussion In this paper we shall discuss lanthanide-induced shift (LIS) data for a series of thietane 1-oxides (1-1 1) in terms of the conformational equilibrium shown in eq 1, where (1)For a review see: Moriarty, R. M. Top. Stereochem. 1974, 8 , 271-421. (2) (a) Johnson, C. R.; Eiegl, W. 0. J . Am. Chem. SOC.1969,91, 2796. (b) Tetrahedron Lett. 196'9, 1879. (c) Siegl, W. 0.;Johnson, C. R. Tetrahedron 1971, 27, 341. (d) J . Org. Chem. 1970, 35, 3657. (3) Barlow, J. H.; Hall, C. R.; Russell, D. R.; Smith, D. J. H. J . Chem. SOC.,Chem. Commun. 1975, 133.

Ln represents the shift reagent Eu(dpm), or Yb(fod)3. s

s20.-.-Ln

b

S

R

t

II

,.o

Rc

L;; ' I

IT

Chloroform solutions of the sulfoxides were doped with increasing amounts of Eu(dpm), or Yb(fod),, and the proton (and in some cases carbon) NMR spectra were recorded as a function of shift reagent concentration. From these data the relative LIS for most protons (and carbons) could be determined (Table I). Previous work4 has shown that the shift reagent coordinates with the sulfoxide oxygen. If the location of the shift reagent relative to the sulfoxide is known, then it should be possible to calculate the relative LIS for comparison with the experimentally determined ones. Good agreement implies that the shift reagent has been properly located, that the substrate's structure is correct, and that the NMR spectrum is properly a ~ s i g n e d . In ~ our study we have assumed that all the thietane 1-oxides have the same shape, namely that reported for thiethane 1-oxide (14).6 This is probably a good (4) (a) Uebel, J. J.; Wing, R. M. J . Am. Chem. SOC.1972,94,8910. (b) Wing, R. M.; Uebel, J. J.; Andersen, K. K. Ibid. 1973, 95, 6046. (5) For a review of the lanthanide-induced-shift technique and its application in conformational analysis see: Hofer, 0. Top. Stereochem. 1976, 9, 111-97.

0022-3263/79/1944-4757$01.00/0 0 1979 American Chemical Society

4758 J . Org. t7hem , Vol. 34, No. 26, 1979

Smith et al.

Table I. Chemical Shifts and Lanthanide-Induced Shifts (LIS) (ppm) for

HC

__________

__

__

_ _ _ _ ~ _ _ _

' H shiftslrelative LIS (Eu( dpm),)

compd ( R J R t ) 1 (CH J C H , ) 2 (HICH 3 ) 3 (CH,/H) 4 (Hit-Bu) 5 (t-BulH) 6 ( HIPh I 7 (PhlH) 8 (Hl4-BrPh) 9 (4-BrPhIH) 10 (PhlCH,) 11 (CH , / P h )

Hc 3.00/100 3.2'IlOO 2.80/100 3.3' 1100 2.92/100 3.62'1100 3.3'1100 3.6'/100 3.2'/100 3.461100d 3.31100

Ht 3.50158 3.2'163 3.65166 3.3'165 3.92164 3.62'174 3.90168 3. 6' / 6 I c 3.86166 3.b6150d 4.0156

1.29127 3.0' 156 1.22127 3. I' 163 0.92124 4.39165

1.22133 1.25132 2.32 149 0.90122 1.93161 7.3117

7.3117

3.3'163

4.38163' 7.08, 7.44124' 7.211. . . 1.68136

7.10, 7.49117'

3.2'159 1.48/25* 7.2113

--___

1/11

10'1R'

100/0 86/14 lOO/O 68132 lOO/O 75/25 lOO/O 74/26 lOO/O 10or/o 10010

0.1 0.4

5.8 6.8

6.0 5.3 5.1 4.5 0.1

6.0' 0.2

C-13 shifts1relative LIS ( Y b ( fod),) 1 (CH,/CH,) 10 (PhlCH,) 11 (CH,/Ph)

63.8/100f 63.31100 62.71100

'

28.616gf 34.4169 37.1181

32.0/34f

29.2137f 30.4136

34.3148

Agreement factor, [ \ ( o b s d calcd)?/l:( o b ~ d ) * ] " ~ . ^ *Estimated " chemical shift. Spectrum unresolved. Shift as measured from centel, of AA'BB' pattern. Yb(fod),-induced shifts. '' Results obtained by combining proton and carbon LIS data. f Yl:( dpm)-induced shifts; private communication from M. R. Willcott. a

approximation since 1,6 8,39,3and 146 all have about the same dihedral ring m~gles-33 f 2'. The location of the lanthanide shift reagent and the ratio (I/II) were treated as unknowns. They therefore were varied to obtain the best agreement7 between calculated and observed LIS. Only chemically reasonable solutions were allowed. The lanthanide was positioned within the OSC(3) plane, syn to sulfur's lone electron pair, a t a distance of 2.5 f 0.5 A from the oxygen, with a Ln-0-S angle of 110-1'70'. lJsing these restrictions, we found (Table I) that all the cis sulfoxides (3, 5, 7, and 9) gave their best correlations u ith the substituents in the diequatorial conformation, Le., conformation I. Attempts to mix in small amounts, of the diaxial conformation always led to considerably poorer results. The best correlations tended to have the lanthanide located 2.0-2.4 A away from the oxygen, with an average Ln-0-S angle of 160 f 10'. Similarly, all three 3,3-disubstituted compounds strongly prefer the equatorial oxygen conformation, i.e., conformation I. Once again the best correlations tended to have the lanthanide located relatively close to the oxygen (2.1-2.5 A), but the average Ln-0-S angle was more acute (120 f 4'). The variations observed in the location of the lanthanide reflect uncertainties in the LIS data, undoped chemical shiftlj, and the geometry of the ring. The more interesting and more difficult questions concern the trans 3-substituted thietane 1-oxides (2, 4 , 6 , and 8). The best correlations for all these compounds tended to have a somewhat longer Ln-0 distance (2.5-3.0 A) with fairly acute Lri-0-S angles (100-125'). These metal locations are agam reasonable. Every trans isomer preferred the equatorial oxygen conformation to various degrees (Table I, compounds 2. 4,6,and 8,or Table 11, method A). The trends shonn in Table I1 are reasonable and are in line with the steric requirements of these groups when they are attached to a cyclohexane ring.8 Plots of the % (6) Bevan, J. W.:Legon, A. C.; Miller, D. J. J . Chem. Soc., Chem. Commun. 1974, 669. (7) Davis, R. E.; Vv'i1lcoi.t 111. M. R. J . Am. Chem. SOC.1972. 94, 1744. ( 8 ) Hirsch, J. A. Top. Stcrcorhem. 1967. I , 199-222.

Table 11. Substituent Effects o n t h e Conformations of t r a m - 3 -Subst it uted-Thietane 1-Oxides % equatorial oxygen conformation ( I ) substituent

method A' 861 2 7 5 - 20 74z 3 6 8 + 20

method Ba

_________________________

CH 3 C,H7 p-BrC,H, t-Bu

' Consult text

87

76 75 74

for explanation.

10

08

06

04

N (equatorial oxygen confwrnation)

Figure 1. Plots of calculated agreement factors (% R ) for trans sulfoxides 2, 4, 6, and 8 as a function of the mole fraction of the equatorial oxygen conformer (I, eq 1).

R factors vs. mole fraction of the equatorial oxygen conformation (Figure 1) show the sensitivity of the correlations to the conformational equilibrium. The methyl and p bromophenyl cases are the most sensitive, and the tertbutyl and phenyl cases are the least. The ranges given

J . Org. Chem., Vol. 44, No. 26, 1979 4759

3-Substituted l'hietane 1-Oxides Scheme I

listed under method B in Table 11. The results of both methods compare quite favorably. These conformational estimates are, of course, valid only for the lanthanide-bound sulfoxides (1.L 1I.L) and therefore they represent an estimate of K,. The extent to which K , the conformational equilibrium constant for the thietane 1-oxides (I F= 11),is perturbed by coordination to the lanthanide shift reagent is under investigation.

3

L.1

Y 1I.L K

1.L 51I.L

Experimental Section

2

VZlH

H

3 61 680(4201 127 31

1'64(1131

H

E

Infrared spectra were recorded on a Perkin-Elmer 237 grating spectrometer or a Perkin-Elmer 580 grating spectrometer. Proton NMR spectra were recorded on a Varian T60 or a JEOL-PS100 spectrometer using deuteriochloroform as solvent and tetramethylsilane (Me4Si) as internal standard. Carbon-13 NMR spectra were recorded on a JEOL FX60 spectrometer operating a t 15 MHz in the F T mode with Me4% as internal standard and deuteriochloroform as solvent. Chloroform solutions of the sulfoxides (ca. 50 mg/mL) were doped with increasing amounts of Eu(dpm), or Yb(fod), and the NMR spectra were recorded as a function of shift reagent concentration up to a molar ratio of about 0.2 (shift reagent/sulfoxide). The relative LIS were determined from the linear plots of the induced shifts vs. amount of added shift reagent. The relative LIS were analyzed in terms of equilibrium 1 with the aid of a modified version of a previously described computer p r ~ g r a m . ~ The thietane 1-oxides (1-1 1) were prepared from the appropriate thiete sulfones.2d Reduction to the thietane sulfones was achieved using sodium borohydridezd or (for 3-tert-butyl-, 3phenyl-, and 3-(p-bromophenyl)thietane sulfones) by catalytic hydrogenation using 10% palladium on barium carbonate. The thietane sulfones were reduced to the sulfides, using lithium aluminum hydride, and reoxidized to the isomeric sulfoxides by stirring overnight at room temperature with an equivalent amount of rn-chloroperoxybenzoic acid dissolved in dichloromethane. Careful chromatography of this mixture of isomers (typically 1.2 g) on silica (50 g) and eluting with ether gave the cis isomer first, closely followed by the trans isomer. Confirmation of the stereochemical assignments of these compounds was based on NMR spectra described previously.2 Preparation of the 3-(p-Bromophenyl)thietane1-Oxides (8 and 9). 3-(p-Bromophenyl)thiete1,l-Dioxide. A mixture of p-bromoacetophenone (80 g, 0.4 mol), morpholine (52.5 g, 0.6 mol), p-toluenesulfonic acid (0.1 g), and benzene (250 mL) was refluxed in a water separator until no more water separated (5 days). The solvent was removed under reduced pressure to give p-bromo-n-morpholinostyrene as a yellow oil. Methanesulfonyl chloride (47.6 g, 0.4 mol) was slowly added to an ice-cooled solution of the crude enamine (0.4 mol) and triethylamine (41.9 g, 0.4 mol) in benzene (250 mL). The mixture was allowed to warm to room temperature and stirred overnight. The product was filtered, washed with water to remove triethylamine hydrochloride, and dried. The crude product was oxidized and the amine oxide was pyrolyzed as described by Johnsonzd for the preparation of 3phenylthiete 1,l-dioxide. Recrystallization of the crude product from ethanol gave 3-(p-bromophenyl)thiete 1,l-dioxide (62 g, 60%) as white crystals: mp 228-230 "C (from ethanol); IR u,, (chloroform) 1600, 1312, 1209, 1130, 785 cm.'; NMR 6 7.53 (2 H, d, J = 9 Hz), 7.20 (2 H, d , J = 9 Hz), 6.87 (1 H, s), 4.70 (2 H, s); MS m / e 260, 258, 212, 210, 196, 194, 183, 181, 131, 115. Anal. Calcd for C9HiBr02S: C, 41.7; H, 2.7; S,1 2 35. Found: C, 41.6; H, 2.75; S, 12.5. 3- (p-Bromophenyl)thietane 1,l-Dioxide. Procedure A. The thiete sulfone above (22 g, 0.09 mol) was added in small portions over 2 h to a mixture of sodium borohydride (15 g, 0.4 mol) and isopropyl alcohol (50 mL) at 60 "C. The mixture was stirred for 4 days. A further 5 g (0.13mol) of sodium borohydride was added after the first 2 days. The mixture was cooled and made slightly acidic by dropwise addition of dilute sulfuric acid. The solvent was removed under reduced pressure and the residue

5)??216(821

3 32(2871 644(426 H1

I6419 I 1

223(3001

-

I3

Figure 2. Proton and carbon chemical shifts (in parts per million downfield from Me,Si) and the relative LIS induced by Yb(fod), (in parentheses) for 12 ,and 13. The proton and carbon LIS are roughly scaled to one another. For 12, using the proton LIS, we calculate r (Yb-0) = 2.0 8, 0 (Yb-0-S) = 120°, and R = 3.4% and, using the carbon LIS, r = 2.0 A, 0 = 130°, and R = 2.7%. For 13, using the proton LIS, we calculate r = 2.3 A, 0 = 161", and R = 2.6% and, using the carbon LIS, r = 2.0 A, 0 = 152", and R = 2.9%.

under method A represent the amount a conformational fraction must be changed to cause the R factor' to increase by 1.7. A change of 1.'7 represents the 95% confidence level for a hypothesis test of order (4,l) using Hamiltong statistics. In method A the lanthanide was positioned so as to obtain the best correlation. The location of the metal was assumed to be the same in both conformations. Models suggested that this was a reasonable approximation. However, in an effort to test this and a t the same time obtain models of equatorial and axial locked thietane 1-oxides, sulfoxides 12 and 13 were prepared (Scheme I). Sulfoxide 13 was previously prepared by Lautenschlaeger'O using the method outlined in Scheme I. The 'H and I3C LIS data (Figure 2) correlate very well with these structures. The lanthanide-oxygen distances are nearly the same (ca. 2.0-2.3 A) for either compound, whereas the Ln-0-43 angle for 12 is somewhat larger (15C-160") than for 13 (12C-130"). Larger angles were also observed for the equatorial locked cis 3-substituted thietane 1-oxides (3, 5, 7 , and 9) and the 3,3-disubstituted compounds (10 and 11). These results support the previous assumption that the lanthanide is similarly located in either conformation. They also suggest an alternative approach to the conformational analysis of the trans sulfoxides. The lanthanide locations found for 12 and 13 could be assumed to hold for the respective conformations of the trans sulfoxides. The conformational percentages which gave the best fits using these metal locations are (9) Hamilton, W.C. Acia Crystallogr. 1965, 18, 502. (10) Lautenschlaeger, F. J . Org. Chem. 1969, 34, 3998.

4760 J . Org. Chem., Vol. 44, No. 26, 1979 was extracted with hot ethyl acetate. Removal of the solvent gave 3-(pbromophenyl)thielane 1,l-dioxide (11g, 50%): mp 154-155 (chloroform) 1495, 1325 (s), 1220, "C (from methanol); I13 u, 1140 (s), 1015,825 cm-':, NMR 6 7.32 (2 H, d, J = 9 Hz), 3.28-4.65 (5 H, m); m/e 262,260,198, 197,196,194,182,133,118,117,116, 115, 103, 102. Anal. Calcd for CgHgBrOzS:C, 41.4; H, 3.45; S, 12.25. Found: C, 41.3; H, 3.5; S, 12.3. Procedure B. The thiete sulfone (5 g, 0.02 mol) and 10% palladium on barium cilrbonate (1g) were stirred in ethanol (50 mL) in an atmosphere of hydrogen overnight. The solution was filtered and the solvent was removed under reduced pressure to give a product identical with the above (4.5 g, 90%). 3-(p-Bromophenyl)thietane. 3-(pBromophenyl)thietane 1,l-dioxide was reduced with lithium aluminum hydride, as described by Johnsonzdfor the preparation of 3-phenylthietane, to give 3-(pbromophenyl)thietane (35%). 3-(p-Bromophenyl)thietane 1-Oxides (8 and 9). A solution of m-chloroperoxybenzoicacid (1.07 g, 5.2 mol) in dichloromethane (30 mL) was slowly added to an ice-cooled solution of 3-(pbromopheny1)thietane (1.1g, 4.8 mmol) in dichloromethane (35 mL). After 3 h the solrution was allowed to warm to room temperature and filtered. The filtrate was washed with aqueous sodium carbonate. Removal of the solvent under reduced pressure gave a clear oil (1g). Chromatography on silica (50 g) and elution with ether-methanol (Ei0:l) gave cis-3-(p-bromophenyl)thietane 1-oxide (0.45 g, 38%): mp 108-109 "C (from chloroform-petroleum ether); IR ifmax (chloroform) 1495,1142,1065 (br), 1012,820 cm-'; IR u, (Nujol) 1490, 1138, 1060 (br), 1012,823 cm-'; NMR 6 7.37 (2 H, d, J = 8.5 Hz), 6.98 (2 H, d, J = 8.5 Hz), 3.17-4.00 (5 H, m); MS m/e 246,244,197,195,184,182,171,169,117,116, 115, 103, 77. And. Calcd for C&J3rOS: C, 44.1; H, 3.65; S, 13.05. Found: C, 43.8; H, 3.65; S, 13.1. trans-3-(p-Bromophenyl)thietane 1-oxide (0.46 g, 39%) was eluted next: mp 112-113 "C (from chloroform-petroleum ether); IR ,v (chloroform) 1495, 1095, 1078, 1060 (br), 1011,827 cm-'; IR urn, (Nujol) 1495,1095,1075, 1059 (br), 1010, 790 crn-l; NMR 6 7.35 (2 H, d, J = 8.5 Hz), 6.98 (2 H, d, J = 8.5 Hz), 4.07-4.2 (1H, m), 3.64 (4 H, d, J = 6 Hz); m/e 246,244,230,228,197,195,184,182,171,169,117,116,115, 103, 102, 91, 77. Anal. Found: C, 44.35; H, 3.8. 3-Methyl-3-phenylthietane 1,l-Dioxide. 2-Methyl-2phenylpropane-l,3-diol was prepared by the Tollens condensation of 2-phenylpropionaldehyde (45.2 g) and 40% aqueous formaldehyde (105 g) in the presence of anhydrous potassium carbonate (33.7 g), ethanol (100 mL), and water (30 mL). The reactants were refluxed for 5 h, tifter which most of the volatile material was removed on a rotary evaporator and the residues were dissolved in dichloromethane. The insoluble materiib were removed by filtration and washed with dichloromethane, after which the organic solutions were combined and the solvent was removed to give a heavy viscous oil (21 g) which solidified on standing to give a white solid. Some of this diol (113.5 g) was dissolved in dry pyridine (110 mL) and stirred at 0 " ( 2 . To this was added dropwise 2.2 equiv of benzenesulfonyl chloride while the temperature was kept below 12 "C with an ice bath. When the addition was complete, the mixture was allowed to warm to room temperature and stirred overnight. The product was isolated by pouring the resultant mixture into water (370 mL) while stirring vigorously. Filtration of the resultant precipitate yielded the crude dibenzenesulfonate ester as a pink solid, which was washed with water, 2 M hydrochloric acid, and water and air dried to give the ester, which was considered pure enough for direct use. Toluene (100 inL) was added to the mixture of crushed sodium sulfide nonahydrate (24.5 g) and dimethyl sulfoxide (200 mL), and the mixture was heated under reflux with a Dean-Stark apparatus collecting thls water which began to appear a t 120 "C. When about 15 mL of water had been azeotroped from the system, the excess toluene was removed under slight vacuum. The resultant Me2S0 solution was slightly green with small gummy residues. The solution was cooled to 90 "C, and some of the previously prepared ditienzenesulfonate ester (30.4 g) was added to the stirred solution over 1 h. The mixture was stirred at 90 "C for 1.5 h more, after which it was cooled and poured into water (1 L). The aqueous solution was extracted with petroleum ether (5 X 100 mL), and the 'combined petroleum ether extracts were washed with water (2 x 100 mL). Drying and removal of the

Smith e t al. solvent gave a yellow oil (15 g), which on distillation afforded 3-phenyl-3-methylthietane (12 g) as a yellow oil: bp 140-160 "C (15 mm); NMR 6 6.82-7.48 (5 H, m), 3.79 (2 H, d, J = 9 Hz), 2.99 (2 H, d, J = 9 Hz), 1.78 (3 H, s). Some of this material (2 g) was oxidized with peracetic acid (10 mL) to give 3-phenyl-3-methylthietane 1,l-dioxide as a white crystalline solid (2.1 g, 84%): mp 48-49 "C (chloroform-petroleum ether); NMR 6 7.12-7.71 ( 5 H, m), 4.39 (4 H, A'B' quartet), 1.82 (3 H, s); IR ,,u (dichloromethane) 1409,1330, 1226,1142,1089 cm-'; m/e 196, 132, 118, 117, 103, 91, 78, 77. Anal. Calcd for C10H12S02:C, 61.20; H, 6.16; S, 16.34. Found: C, 61.32; H, 6.36; S, 16.30. cis-and trans-3-Phenyl-3-methylthietane 1-Oxide (10 and 11). 3-Phenyl-3-methylthietane (1g) was oxidized to the isomeric sulfoxide using m-chloroperoxybenzoic acid as described for the previous examples. Careful chromatography of this mixture on silica (50 g) using ether as the eluent gave the cis isomer as a clear oil (632 mg) followed by the trans isomer as a white crystalline solid (SO0 mg). Preparation of exo-8-Thiatetracyclo[2.2.l.l.0]octane 8Oxidelo (13). The addition product of sulfur dichloride and norbornadiene (20 g) was treated with hot aqueous sodium carbonate. The reaction was worked up after 30 min; there was no change. Treatment for 3 h, however, gave a small amount of the desired sulfoxide after workup (1.93 g, 13.4%): mp 72-78 "C [lit.lo 76-77 "C]; NMR 6 4.31-4.55 (1 H, br m), 3.23-3.44 (2 H, m), 1.98-2.22 (1H, m), 2.53-2.90 (4 H, m); IR urn= (dichloromethane) 3050,2939,2862,1125,1049 cm-'; MS m/e 140,123,91; 13CNMR (ppm downfield of Me,Si) 9.82, 21.18, 34.09, 45.13, 64.36. Preparation of endo-8-Thiatetracyclo[2.2.l.l.0]octane %Oxide(12). Some of the previously prepared exo sulfoxide (1.2 g) was suspended-dissolved in dry ether (50 mL) with stirring a t 0 "C. Lithium aluminum hydride (0.2 g) was slowly added to this mixture over 10 min with rapid stirring. The reaction was allowed to warm to room temperature and stirred for 4 h. Usual workup gave a bad-smelling colorless oil. The NMR spectrum still showed some traces of the starting material but there were other resonances at 6 3.53-3.72 (m), 2.89-3.08 (m), and 2.15-2.49 which could be reasonably assigned to the sulfide. TLC on silica eluting with ether showed two spots, one of them coincident with the starting material, Rf0.052, and the other one Rf0.84. This material was stirred in excess aqueous sodium metaperiodate (3 g unsaturated) for 2 days a t room temperature, after which extraction with dichloromethane and normal workup gave a white solid (1.0 g). The NMR spectrum showed none of the exo sulfoxide. TLC (silica) eluting with ether showed two spots, Rf 0.032 and 0.26. Careful chromatography of some of this material (0.75 g) on silica (80 g) eluting with pure ether increasing to 2% methanol in ether gave two compounds as white crystalline solids. The first compound to be eluted (R, 0.26) was 8-thiatetracyclo[2.2.l.l.O]octane 8,8-dioxide (350 mg): mp 135 "C dec (lit.'O 130 "C dec); NMR 6 3.90-4.10 (2 H, m), 2.70-2.95 (1 H, br m), 2.10-2.37 (1 H, m), 1.87-2.10 and 1.66-1.85 (4 H, m); IR ,u (dichloromethane) 3060, 2950, 2815, 1300, 1175, 1097; MS m/e (no molecular ion) 92, 91, 65, 65; 13C NMR (ppm) 13.23, 21.34, 31.17,34.90,76.78. The second compound, R, 0.032 (101 mg), gave the following data: mp 113-115 "C; NMR 6 3.49-3.79 (2 H, m); IR u- (dichloromethane) 3051,2942,2875, 1138,1080 cm-'; m/e 140,123,91; '% NMR (ppm) 21.02,27.35,32.55,36.36,68.01.Anal. Calcd for C7H80S: C, 59.99; H , 5.75; S, 22.84. Found: C, 60.06; H, 5.76; S, 22.54. The remainder of the crude oxidation mixture (250 mg) was dissolved in peracetic acid (4 mL) and stirred for 4 days. Workup gave a white solid (200 mg) with the same NMR spectrum as that of the known sulfone."

Acknowledgment. We a r e grateful to the Eastman Kodak C o m p a n y for its assistance with t h e preparation of the manuscript and its hospitality to J.J.U. during his sabbatical leave.

Registry No. 1, 27832-55-9; 2, 25145-03-3; 3, 25145-02-2; 4, 25902-69-6;5, 25902-66-3;6, 26091-84-9;7, 25902-67-4;8, 55779-41-4; 9, 55779-40-3;10, 66809-92-5;11, 66810-23-9;3-methylthiete 1,l-dioxide, 71887-62-2; 3-tert-butylthiete 1,l-dioxide, 25903-16-6; 3phenylthiete 1,l-dioxide,25903-17-7;3-(pbromophenyl)thiete 1,l-

J . Org. Chem., Vol. 44, Y o . 26, 1979 4761

Transannular Interactions in Cage Compounds dioxide, 71887-63-3; 3,3-dimethylthietane 1,l-dioxide, 27832-56-0; 3-methylthietane 1,l-dioxide, 25903-07-5;3-tert-butylthietane 1,ldioxide, 25903-05-3;3-phenylthietane 1,l-dioxide, 25636-64-0;3-(pbromopheny1)thietane 1,l-dioxide,71887-64-4;3-methyl-3-phenylthietane 1,l-dioxide,66809-99-2;3,3-dimethylthietane, 13188-85-7; 3-methylthietane. 22438-40-0;3-tert-butylthietane, 25903-02-0; 3phenylthietane, 25636-63-9;3-(p-bromophenyl)thietane,55779-42-5;

3-methyl-3-phenylthietane, 66810-25-1;p-bromoacetophenone, 9990-1; morpholine, 110-91-8;p-bromo-n-morpholinostyrene,5577943-6; 2-methyl-2-phenylpropane-1,3-diol, 24765-53-5; 2-phenylpropionaldehyde, 93-53-8; 2-methyl-2-phenylpropane-1,3-diol di-

benzenesulfonate ester, 66810-41-1; norbornadiene, 121-46-0; 8thiatetracyclo[2.2.1.l.O]octane 8,8-dioxide,22061-75-2;12, 71927-861; 13, 22061-74-1.

Correlation of Structure and Transannular Interactions in a Series of Cage Compounds J a m e s C. Barborak,*la Diana Khoury,Ib Wilhelm F. Maier,Ic P a u l v. R. Schleyer,l‘ E d w a r d C. Smith,Ia William F. S m i t h , Jr.,la and Christopher WyrickIa Departments of Chemistry, University of North Carolina at Greensboro, Greensboro, North Carolina 27412, and Princeton Cniversity, Princeton, Neu Jersey 08540, and the Institut fur Organische Chemie, Universitat Erlangen-Niirnberg, 8520 Erlangen, West Germany Received October 13, 1978 A series of cage systems with general structure 7 were synthesized. By variation of bridge C, the distance between functional groups A and B could be changed systematically and the chemical consequences of these changes investigated. Thus, Diels-Alder reactions between benzoquinone and cyclic dienes ranging from cyclobutadiene to cycloheptadiene, and including cyclooctatetraene, yielded endo dienes which were subsequently converted to cage structures photochemically. This transformation results in compounds 8, 10, 15, 16, and 17 with two transannular carbonyl groups positioned a t varying distances. Under ketalization conditions some of these compounds--the ones with larger bridges C which brought the keto groups closer togetheygave cyclized products rather than simple ketals. In the other cases, monoketals formed which were reduced to endo alcohols and studied further. Molecular mechanics calculations on hydrocarbon models for these systems provided a more quantitative basis for interpretation of the experimental results. The distance between A and B in 7 varied with the size 39) became more of bridge C. The calculated energies of simulated ring-closure reactions (40 39 and 41 favorable as the number of atoms in bridge C was increased. These results parallel the experimental trends.

-

Rigid cage molecules a r e valuable substrates for t h e s t u d y of organic chemical transformations.2 Compounds of this type have marked advantages over conformationally mobile molecules. Reaction centers in rigid molecules are fixed with respect t o the remainder of t h e molecular skeleton; hence, p e r t u r b a t i o n s d u e t o conformational changes are effectively diminished or removed. T h i s greatly simplifies the understanding of m a n y chemical transformations a n d permits analysis of structure-reactivity relationships. Not only mechanistic studies but also organic synthesis has t a k e n advantage of t h e potential interaction between t w o reactive centers in conformationally restricted molecules. Polycyclic c o m p o u n d s of great rigidity and often of high symmetry, the so-called cage molecules, can result from transannular ring closure. For example, adamantanes and closely related compounds have been prepared from semirigid precursors. T h u s , acetolysis of t h e spirocyclic tosylate 1 gives t h e homoadamantyl acetate 2,3 and 7 methylenebicyclo[3.3.l]nonan-3-one3 provides t h e noradamantane 4 by red.uctive ~ y c l i z a t i o n . ~T r a n s a n n u l a r cyclizations of compounds such as 5 to produce the bridged s t r u c t u r e 65 a r e germane t o t h e present work. We wished t o s t u d y t h e variations of chemical interactions in a series of c l o d y related rigid molecules in which

1

-

3

2

7

4

5

6

t h e transannular distance between functional groups could b e altered systematically. The generalized pentacyclic system 7 fulfills these requirements. C o m p o u n d s of t h i s

(1) (a) University of North Carolina at Greensboro. (b) Princeton

University. (c) Universitat Erlangen-Numberg.

7

(2) For reviews see: Bingham, R. C.; Schleyer, P. v. R. Fortschr. Chem. Forschung. 1971,18, 1-102. McKervey, M. A. Chem. SOC.Reu. 1974.3, 479-512. Fort, R. C., Jr. “Adamantane. The Chemistry of Diamond Molecules”; Marcel Dekkei-, Inc.: New York, 1976. (3) Eakin, M. A ; Martin, J . C.; Parker, W.Chrrn. Cornrnun. 1967,

kind a r e readily available b y sequences of Diels-Alder reactions a n d photochemical ring closures. The bridge C (in s t r u c t u r e 7) c a n be modified in order to alter t h e transannular relationship of functional groups A and B. Possibilities for such modifications a n d their consequences have been recognized previously.6 Empirical force field

955-6. (4) Eakin, M. A,; Martin, J. C.; Parker, W. Chem. Commun. 1965. 206. (5) Sasaki, T.; Eguchi, S.; Kiriyama, T.; Hiroaki. 0. Tetrahedron 1971, 30, 2707.12.

0022-3263/79/1944-4761~01.00/0

c

1979 American Chemical Society